What is an ATS and why does every serious employer use one?
An applicant tracking system (ATS) is software that sits between your submitted CV and the hiring team's inbox. When a company posts a job and receives applications, the ATS ingests those applications, parses their content into structured fields, scores them against criteria set by the recruiter, and presents ranked candidates to the hiring team. For popular roles at mid-to-large companies, this means your CV is evaluated by software before any human reads it.
The business logic is straightforward: a marketing director role at a mid-size company might receive 400–800 applications within 72 hours of posting. Without software to triage that volume, hiring teams would collapse under the administrative load. The ATS culls that pool to a shortlist of 20–50 candidates that get human attention. If your CV doesn't survive that cull, you never get considered — regardless of how qualified you actually are.
Well-known ATS platforms include Greenhouse, Lever, Workday, iCIMS, Taleo, Jobvite, and Bullhorn. Each has its own parsing and scoring logic, but they share common principles. Understanding those principles — rather than trying to game any one platform — is the key to writing CVs that consistently get through.
How ATS parsing actually works
When your CV arrives, the ATS's parser attempts to convert it from an unstructured document into structured database fields: name, contact information, work experience entries (each with company name, job title, dates, and bullet points), education entries, and skills. This parsing is largely pattern-matching against expected document structures.
Most modern parsers are reasonably good at handling standard CV structures in common file formats. Where they fail consistently is unusual formatting: content inside tables, text boxes, headers and footers, columns created with tab characters rather than actual column layout, and inline images or logos. These elements either parse incorrectly (appearing in the wrong field) or don't parse at all (the content disappears). A recruiter who opens your ATS profile might see a correctly formatted experience entry in what you intended as column two — now floating at the bottom of an otherwise empty candidate record.
Parsing also has implications for dates and job titles. ATS systems typically identify employment periods by looking for date patterns adjacent to company names and job titles. Non-standard date formats (using words like "Summer 2022" or "Q3 2021" instead of month-year) can confuse parsers and cause your experience to either be incorrectly dated or dropped entirely. Use clear, consistent date formats: "Jan 2021 — Mar 2023" or "2021–2023" are both reliably parsed.
Why qualified candidates get auto-rejected
The most common reason qualified candidates get rejected by ATS before a human sees them is a mismatch between the language in their CV and the language in the job description. A recruiter builds an ATS search or scoring filter using specific terms: "Python," "project management," "stakeholder engagement," "Series A." A candidate whose CV says "scripting in Python," "programme management," and "executive alignment" may be a perfect fit — but if their exact phrasing doesn't match the search terms, their score is suppressed and they don't make the shortlist.
This isn't a flaw in the technology; it reflects the reality that language for the same work varies considerably across industries, company sizes, and geographies. An experienced hire who has always worked in one sector may be genuinely qualified for a role in an adjacent sector but consistently loses to less-experienced candidates from within the sector who naturally use the target vocabulary.
The second major cause of auto-rejection is formatting issues that cause the ATS to misparse the CV, making it appear less qualified than it is. If your most recent employer's name is in a text box that the parser can't read, your experience starts with a role from 2019. The ATS scores you as having a 6-year gap and deprioritises you automatically, even though your profile is actually excellent.
Formatting rules that guarantee clean parsing
The safest CV format for ATS submission is a single-column document using standard heading styles, plain paragraph text for bullet points, and no graphical elements. This is not a concession to ugliness — a well-designed single-column CV can look professional and distinctive. It is, however, a concession to reliability: these documents parse cleanly across all major ATS platforms without exception.
Specifically, avoid the following: tables used for layout purposes (fine for data, problematic for CV structure), text boxes and shapes, headers and footers that contain important information like your name or contact details (the parser often ignores these zones), multi-column layouts created with columns in Word or Google Docs, and inline images including logos, headshots, or charts. Also avoid "fancy" fonts — while they display correctly in the PDF, ATS OCR processing of PDFs can misread unusual glyphs.
For bullet characters, use standard ASCII characters (the hyphen - or the bullet •) rather than decorative symbols that Unicode encodes differently. For section headings, use consistent naming that ATS systems recognise: "Work Experience" or "Professional Experience" rather than "Where I've Been" or "Career Story." Creative heading names that would charm a human recruiter often confuse the parser's section-detection logic.
Keyword strategy: density, placement, and natural language
The most effective keyword strategy is not about hitting a target count — it's about ensuring that your most relevant experiences are described using the vocabulary the job description uses. For each high-priority keyword in the posting, you want at minimum one strong sentence in your CV that uses that keyword in a meaningful context.
Keyword density matters because ATS scoring is often based on weighted counts — how many times a keyword appears, and in which sections. A keyword that appears once in a skills list and twice in experience entries scores higher than one that appears only in the skills list. Context multiplies value. That said, forcing a keyword into every possible sentence quickly makes your CV unreadable and can trigger spam filters in more sophisticated systems. Aim for natural, professional prose that happens to use the right vocabulary.
The professional summary is a high-value zone for primary keywords because it's typically weighted more heavily than body content in ATS scoring. If the job description prominently features "product strategy" and "cross-functional leadership," your summary should contain both — naturally, not mechanically. Experience bullet points are the bulk of your keyword real estate. Education and skills sections provide supplementary signal but are typically weighted lower than work experience content.
File format advice: PDF vs DOCX
The age-old debate about whether to submit your CV as a PDF or DOCX has a nuanced answer that depends on the specific ATS. PDFs preserve your formatting perfectly across all devices and operating systems, which is why most candidates prefer them. However, some older ATS platforms (particularly Taleo, which is still widely deployed) have historically parsed PDFs less accurately than DOCX files. They either run the PDF through an OCR engine or attempt to extract text from the PDF's content stream — both of which can produce errors with complex layouts.
Modern ATS platforms — Greenhouse, Lever, Workday (newer versions), iCIMS — have substantially improved PDF parsing and generally handle clean, text-based PDFs well. The key phrase is "text-based PDF": a PDF created by saving a Word or Google Docs document as PDF, where the text is stored as actual characters, parses cleanly. A PDF that is a scan or photo of a document (even a high-resolution one) will not be parsed reliably by any ATS — it requires OCR and introduces errors.
The safest practical approach: submit as a simple, single-column PDF when the application portal doesn't specify a preference. If the job listing or portal explicitly asks for DOCX, use DOCX. If you're applying through a highly specialised platform for older enterprises (some government contractors, some large manufacturers), having a clean DOCX backup version is worthwhile. Never submit a CV as a JPEG, PNG, or other image format.
Testing your CV before submitting
Before submitting your CV for competitive roles, it's worth running a basic parse test to see what an ATS extracts from your document. Several free and paid tools exist for this purpose — Jobscan, Resume Worded, and RoleFitCV all provide some form of ATS compatibility feedback. These tools simulate the parsing process and show you what fields get extracted, which keywords are detected, and where formatting issues might be causing content to be missed.
A simple DIY test: paste your CV text into a plain text editor like Notepad or TextEdit (with rich text turned off). Everything that appears clearly in plain text will be readable by a text-based ATS parser. Content that was in a text box, table, or header/footer will be missing or jumbled. This test won't replicate the full ATS experience — it doesn't score for keywords or simulate field mapping — but it quickly reveals structural issues.
Finally, before every application, re-read both the job description and your tailored CV side by side. Ask: do the key phrases from the job description appear in my CV? Is my most relevant experience visible in the first third of the document? Does my summary name the specific type of role and sector this posting is for? If yes to all three, your CV is as ATS-ready as it can be — from there, the quality of your underlying experience and the strength of your application letter carry the rest.